SRX Services Gateway
Reply
Regular Visitor
AlexK
Posts: 9
Registered: ‎03-02-2009
0

SRX Mixed Mode (or Six Pack) Cluster and BGP routing

Hi All.

 

Now I'm looking for documentation about deploing SRX Mixed Mode or Six Pack cluster with two BGP uplinks.
I doubt whether I should configure both BGP sessions in global cluster configuration, or I should configure first BGP session in node 0 configuration ( i.e. in groups node0 protocols bgp) and second BGP session in node 1 configuration ( i.e. in groups node1 protocols bgp) ?



 

Regular Visitor
dabernie
Posts: 9
Registered: ‎11-09-2009
0

Re: SRX Mixed Mode (or Six Pack) Cluster and BGP routing

Hi,

 

Based on the diagram you provided, you should configure bgp peers under "protocols bgp".

 

set protocols bgp group node0-peer neighbor x.x.x.x

set protocols bgp group node1-peer neighbor y.y.y.y

 

If you would configure bgp peers under the device specific configuration groups (groups node0 and node1), only the peer configured on the node where RG0 is active (ie the one running the RPD process) would be working. So if RG0 is primary on node0, then only the peer configured under [groups node0 protocols bgp] would be working.

 

And if for any reason RG0 would failover to the other node, than all of a sudden, the active bgp session would fail and the one configured on [groups node1 protocols bgp] would kick in.

 

Thanks

 

Contributor
frank342767
Posts: 21
Registered: ‎03-23-2010
0

Re: SRX Mixed Mode (or Six Pack) Cluster and BGP routing

would you be willing to share your config, I am trying to create a very similar configuration.

Contributor
ed_gpc
Posts: 194
Registered: ‎09-21-2010
0

Re: SRX Mixed Mode (or Six Pack) Cluster and BGP routing

Are you doing active/active with these?

 

Let me know, if so I can shed some light I believe, I just built up an active/active with BGP.

 

Short version though - we created a VR for each public side and then applied the BGP in each VR.  That way in failure the VR "moves" from device A to device B

Copyright© 1999-2013 Juniper Networks, Inc. All rights reserved.